FelixYBW opened a new issue, #7612: URL: https://github.com/apache/incubator-gluten/issues/7612
### Backend VL (Velox) ### Bug description I met a strange issue when I test TPCDS workload. Recode here in case someone also meet. Maybe it's a well-known issue but I spent much time to debug it. It's not related to Gluten but run spark workload on Yarn. I have a master + 1 x worker. worker is configured as ip address. When yarn starts nodemanager on worker, it uses hostname instead of IP address. While the hostname is configured as 127.0.1.1 in /etc/hosts. Then the blockmanager is listening on 127.0.1.1:port. TPCH can pass without any issue, but TPCDS has lots of "connection refused" error, but the query can finish normally in spark-shell. In Pyspark it's hang. The solution is simple, delete the 127.0.1.1 line in /etc/hosts. I remember a configure in Yarn not to resolve hostname but use ip address directly, but I can't find it anymore. ### Spark version None ### Spark configurations _No response_ ### System information _No response_ ### Relevant logs _No response_ -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected] --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
